MAJOR ARTICLE

Evidence of a Novel Human Coronavirus That

Is Associated with Respiratory Tract Disease
in Infants and Young Children
Frank Esper,1 Carla Weibel,1 David Ferguson,2 Marie L. Landry,2 and Jeffrey S. Kahn1,3
1

Department of Pediatrics, Division of Infectious Diseases, and Departments of 2Laboratory Medicine and 3Epidemiology and Public Health,
Yale University School of Medicine, New Haven, Connecticut

(See the brief report by Esper et al., on pages 499502, and the editorial commentary by McIntosh, on pages 48991.)

Infections of the respiratory tract are a leading cause of

morbidity and mortality worldwide [1]. However, for a
substantial proportion of these infections, the agents that
are causing the disease remain undetermined [24]. Recently, there has been renewed interest in human coronaviruses (HCoVs), in part because of the identification of a novel CoV associated with severe acute
respiratory syndrome (SARS) [57]. Before the emer-

Received 7 September 2004; accepted 8 November 2004; electronically published 14 January 2005.
Financial support: Patrick and Catherine Weldon Donaghue Medical Research
Foundation (to J.S.K.); Friends of Yale Pediatrics (to J.S.K.); Yale Childrens Clinical
Research Center (grant M01-RR06022 to J.S.K.); General Clinical Research Centers
Program, National Center for Research Resources, National Institutes of Health
(NIH); NIH (grant T32 AI07210-20 to F.E.).
Reprints or correspondence: Dr. Jeffrey S. Kahn, Dept. of Pediatrics, Div. of
Infectious Diseases, Yale University School of Medicine, P.O. Box 208064, New
Haven, CT 06520-8064 (jeffrey.kahn@yale.edu).
The Journal of Infectious Diseases 2005; 191:4928
 2005 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2005/19104-0002$15.00

492 JID 2005:191 (15 February) Esper et al.

gence of SARS-CoV, HCoVs were generally thought to

cause mild, self-limited infections of the upper respiratory tract [8]. However, it is now known that HCoVs
can cause severe disease in immunocompromised individuals [9]. The study of HCoVs has been hampered
by the difficulty in propagating these viruses in vitro
and the lack of specific diagnostic tests with which to
identify potentially novel viruses. Therefore, the possibility exists that unidentified HCoVs are the cause of
some proportion of the respiratory tract diseases for
which etiological agents have not been identified.
To determine whether novel HCoVs are circulating
and, if so, whether they are responsible for respiratory
tract disease in children, we developed a strategy to
screen for previously unknown HCoVs. Our initial assumption was that all CoVs must have conserved functions and that these conserved functions are reflected
in the genome. The replicase of CoVs is an RNA-dependent RNA polymerase, the function of which is not
provided by the host cell and, therefore, must be evolu-

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

Background. The etiological agents responsible for a substantial proportion of respiratory tract diseases have
not been identified. We sought to determine whether novel human coronaviruses (HCoVs) are circulating in New
Haven, Connecticut, and, if so, whether they are associated with respiratory tract disease in infants and young
children.
Methods. We developed a polymerase chain reaction (PCR)based approach for screening specimens from
the respiratory tracts of symptomatic children. PCR probes that target regions of the replicase 1a gene that are
conserved among genetically diverse animal CoVs and HCoVs were designed. Using these probes, we identified
genomic sequences of a novel HCoV, designated New Haven coronavirus (HCoV-NH). Thereafter, we designed
specific probes to screen respiratory specimens from children !5 years old for this novel HCoV. Clinical features
associated with HCoV-NH infection were identified.
Results. Seventy-nine (8.8%) of 895 children tested positive for HCoV-NH. Cough, rhinorrhea, tachypnea,
fever, abnormal breath sounds, and hypoxia were the most common findings associated with HCoV-NH infection.
Sequence analysis revealed that HCoV-NH is closely related to a novel HCoV recently reported in The Netherlands.
Conclusions. The novel HCoVs identified in New Haven and The Netherlands are similar and likely represent
the same species. This newly discovered virus may have worldwide distribution and may account for a significant
proportion of respiratory tract disease in infants and young children.

PATIENTS, MATERIALS, AND METHODS

Primer design and reverse transcriptase (RT)PCR screening.
Primers for the detection of CoVs were based on conserved
regions of the replicase 1a gene of groups 1, 2, and 3 CoVs
and SARS-CoV. The replicase 1a genes from avian infectious
bronchitis virus (GenBank accession number AJ311317), bovine CoV (BCoV) strain LUN (GenBank accession number
AF391542), BCoV strain Quebec (GenBank accession number
AF220295), BCoV isolate BCoV-ENT (GenBank accession number AF391541), HCoV-229E (GenBank accession number NC_
002645), TGEV (GenBank accession number NC_002306), and
SARS-CoV TOR2 (GenBank accession number AY274119) were
aligned by use of Clustal W in the Lasergene software package
(version 5.05; DNASTAR). Two conserved regions were identified, and primers corresponding to these regions were synthesized (Yale Oligonucleotide Laboratory, Department of
Pathology, New Haven, CT). The forward primer, 5
-GCGCAAAATAATGAATTAATGCC-3
(underscoring indicates a G/C
clamp), and the reverse primer, 5
-GACGCACCACCATATGAATCCTG-3
, represent consensus sequences of conserved regions within the 3
1000 bases of the replicase 1a gene of all of
the CoVs listed above (relative to sequences 11,78112,285 of
HCoV-229E). The predicted length of the amplicons produced
by these primers is 550 bp. RNA from respiratory specimens
obtained from the Clinical Virology Laboratory, YaleNew Haven Hospital, was extracted by use of the QIAamp Viral RNA
Mini Kit (Qiagen), in accordance with the manufacturers instructions. cDNA for each respiratory specimen was synthesized
with random hexamer primers and MuMLV RT (New England
Biolabs), in accordance with the manufacturers instructions.
cDNAs were subsequently screened by PCR with HotStar Taq

polymerase (Qiagen), in accordance with the manufacturers

instructions. The following amplification program was used:
15 min at 95C; 40 cycles of 1 min at 94C, 1 min at 50C,
and 1 min at 72C; and a final extension of 10 min at 72C.
For the initial screening of respiratory specimens, RNA was
extracted from MRC-5 cells infected with HCoV-229E (ATCC
VR-740) as a positive control. Each set of RT-PCR amplifications included appropriate negative controls. PCR products
were analyzed by agarose gel electrophoresis. After the initial
screening of respiratory specimens, amplicons of the predicted
molecular weight were isolated and sequenced. Sequencing was
performed by use of Applied Biosystems 377 DNA automated
sequencers (W. M. Keck Biotechnology Resource Lab, Yale University School of Medicine). Sequences that corresponded to a
potential novel HCoV were identified, and primers specific for
this agent were synthesized. The forward primer, 5
-GCGCTATGAGGGTGGTTGTAAC-3
, and the reverse primer, 5
-CGCGCAGTTAAAAGTCCAGAATTAAC-3
, amplify a 215-bp region
of the novel HCoV genome. These primers target regions of
the novel HCoV genome that are distinct relative to the corresponding region of the HCoV-229E genome. Screening by
PCR with these primers was performed by use of the following
amplification program: 15 min at 95C; 40 cycles of 1 min at
94C, 1 min at 55C, and 1 min at 72C; and a final extension
of 10 min at 72C. These primers were used to screen pooled
respiratory specimens.
Respiratory specimens. We chose to screen specimens that
were obtained from the respiratory tracts of symptomatic children !5 years old and that tested negative for respiratory syncytial virus (RSV), influenza viruses A and B, human parainfluenza viruses 13, and adenovirus by direct fluorescent
antibody assay (DFA). These specimens, which were obtained
from both ambulatory and hospitalized patients, were screened
for presence of human metapneumovirus (hMPV) by use of
an RT-PCRbased approach described elsewhere [12, 13] and
were submitted to the Clinical Virology Laboratory, YaleNew
Haven Hospital, for diagnostic testing. Specimens were obtained
from 1 January 2002 to 14 February 2003. Children could be
counted more than once if the specimens were obtained 130
days apart. Clinical data from the children who tested positive
for HCoV-NH were obtained by extracting information from
medical records and recording the data on a standardized form.
Children who had evidence of infection with another viral respiratory pathogen were excluded when the clinical features associated with HCoV-NH infection were tabulated. Comorbidity
was defined as prematurity (gestational age of !35 weeks), underlying pulmonary disease, genetic syndromes, acquired immunosuppression, malignancies, and congenital heart disease.
The Yale University Human Investigations Committee approved
the collection and screening of respiratory specimens.

Novel Human Coronavirus JID 2005:191 (15 February) 493

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

tionarily maintained by all CoVs. Therefore, we designed polymerase chain reaction (PCR) probes that target regions of the
CoV replicase 1a gene that are conserved among HCoVs, avian
CoVs, and mammalian CoVs. Using this approach, we identified genetic evidence of a novel HCoV circulating in New
Haven, CT. Probes specific for this HCoV, designated New
Haven coronavirus (HCoV-NH), were then used to screen
specimens from the respiratory tracts of symptomatic children.
During our screening of respiratory specimens for this HCoV,
2 studies from The Netherlands reported the identification of
a novel HCoV [10, 11]. This virus is a group 1 CoV and is
most closely related to HCoV-229E and transmissible gastroenteritis virus (TGEV), a virus of pigs. Comparisons of the
sequences of the HCoV identified in The Netherlands and
HCoV-NH revealed that these viruses are closely related and
likely represent the same species. Here, we describe the seasonal distribution and clinical manifestation of disease associated
with HCoV-NH infection.

Sequencing and phylogenetic analysis. The amplicon of each

specimen that tested positive by RT-PCR was sequenced to
confirm the presence of HCoV-NH. Phylogenetic analysis was
performed by use of the Lasergene software package described
above.
RESULTS
Identification of a novel CoV sequence. Using the primers
that target the conserved regions of the CoV replicase 1a gene,
we screened 601 respiratory specimens for CoVs. The screening
reaction was performed on pooled RNA; each pooled amplification reaction included 510 individual specimens. Of the
80 pooled amplification reactions, 17 yielded an amplicon of
550 bp (data not shown). After these amplicons were sequenced, a nucleotide BLAST (available at: http://www.ncbi
.nlm.nih.gov/BLAST/) search was performed. HCoV-OC43 was
identified in 8 pooled reaction products, and HCoV-229E was
identified in 1 pooled reaction product. Six amplicons either
were human DNA or did not yield an interpretable sequence.
The 2 remaining amplicons were similar and represented novel
sequences most closely related to group 1 CoVs. These sequences were 69%71% identical to HCoV-229E and TGEV
on the nucleotide level and 68% identical to HCoV-229E and
TGEV on the amino acid level. PCR primers specific for these
novel sequences of HCoV-NH were then synthesized, and these
primers were used to screen all respiratory specimens thereafter.
Screening of respiratory specimens for HCoV-NH. In total,
1265 respiratory specimens from 895 children were screened
by RT-PCR with primers specific for HCoV-NH (the 601 spec494 JID 2005:191 (15 February) Esper et al.

imens from the initial screening were rescreened by RT-PCR

with these primers). Screening reactions were again pooled,
with 510 individual specimens in each pool. If a pooled reaction produced an appropriately sized amplicon, the specimens in that pool were individually tested. Seventy-nine (8.8%)
of these individual specimens were found to be positive for
HCoV-NH. Two children had 2 positive specimens each; the
specimens were obtained 5 days apart for 1 child and 7 days
apart for the other child. In both instances, the 2 positive
specimens were considered to be the result of a single episode
of HCoV-NH infection. The median age of the 79 HCoV-NH
positive children was 6.5 months. Fifty (63.3%) were !1 year
old, and 27 (34.2%) were 03 months old. Forty-nine (62.0%)
were male. Eleven of the HCoV-NHpositive children had been
hospitalized since birth at the newborn intensive care unit
(NICU), YaleNew Haven Hospital. The age distribution of the
HCoV-NHpositive children is shown in figure 1.
Clinical features associated with HCoV-NH infection.
Clinical data were available for 76 of the 79 HCoV-NHpositive
children. Of these 76 children, 9 (11.8%) had evidence of recent
infection with another respiratory virus2 were coinfected
with hMPV, and 7 were found to be coinfected with another
respiratory virus by DFA on other respiratory specimens that
had been collected during the same hospitalization or illness
(1 had a parainfluenza infection, and 6 had an RSV infection).
Among the 67 children who tested positive for HCoV-NH only,
cough, rhinorrhea, tachypnea, fever, abnormal breath sounds,
and hypoxia were the most common findings (table 1). Thirtyfive (52.2%) of these 67 children had an underlying comor-

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

Figure 1. Age distribution of New Haven coronavirus (HCoV-NH)positive children !5 years old. HCoV-NH infection in children hospitalized since
birth at the newborn intensive care unit (NICU), YaleNew Haven Hospital (CT), is represented by black bars.

Table 1. Clinical features associated with New Haven coronavirus infection in children !5
years old.

Clinical feature

NICU children Non-NICU children

(n p 11)
(n p 56)

Cough
Rhinorrhea
Tachypneaa
Feverb
Rhonchi, rales, or abnormal breath sounds
Hypoxiac
Chest retractions
Wheezing
Stridor
Abnormal chest-radiograph findingsd
NOTE.
a
b
c

(0)
(27.3)
(81.8)
(18.2)
(54.5)
(81.8)
(27.3)
(9.1)
(0)
(71.4)

43
38
30
30
24
16
19
20
4
20/31

(76.8)
(67.9)
(53.6)
(53.6)
(42.9)
(28.6)
(33.9)
(35.7)
(7.1)
(64.5)

43
41
39
32
30
25
22
21
4
25/38

(64.2)
(61.2)
(58.2)
(47.8)
(44.8)
(37.3)
(32.8)
(31.3)
(6.0)
(65.8)

Data are no. (%) of children, except where noted. NICU, newborn intensive care unit.

On the basis of normal values for age-specific respiratory rates.

Temperature 138C.
Oxygen saturation 90%.
Data are no. (%) of abnormal chest-radiograph findings per the no. of chest radiographs obtained.

bidity (19 [28.4%] were born prematurely). Of the 38 children

for whom a chest radiograph was obtained, 25 (66.0%) had
abnormal findings, which were characterized by peribronchial
cuffing, atelectasis, and/or infiltrates.
The clinical features associated with HCoV-NH infection
among the 11 children hospitalized since birth at the NICU
are shown in table 1. The median age of these children at the
time their positive specimens were obtained was 26 days (range,
1151 days). The age distribution and the time of infection of
the HCoV-NHpositive children hospitalized since birth at the
NICU are shown in figures 1 and 2, respectively. Five of the
11 children were infected during February 2002. The hospitalization of all of these children overlapped. Three of the 11
children tested positive for HCoV-NH during a 3-week period
spanning late January and early February 2003; these 3 children
were hospitalized at the NICU at the same time.
Two of the 79 HCoV-NHpositive children died; both had
been hospitalized since birth at the NICU. One child had been
diagnosed with hydrops fetalis in utero and died on day 2 of
life, 1 day after a respiratory specimen had been collected that
tested positive for HCoV-NH. The second child, who had been
born at 28 weeks gestation, died on day 170 of life; a respiratory
specimen collected on day 151 of life tested positive for HCoVNH. This child also had a history of necrotizing enterocolitis
and liver failure before day 151 of life and ultimately died of
multiorgan system failure.
Seasonal distribution of HCoV-NH infection. The weekly
distribution of HCoV-NHpositive children is shown in figure
2, and the percentage of positive specimens detected per month
is shown in table 2. Of the specimens obtained from 1 January
2002 to 31 December 2002, a majority (42/67 [62.7%]) that
tested positive were obtained during the first 10 weeks of the

year. Relatively few positive specimens (13/67 [19.4%]) were

obtained from June to late November.
Phylogenetic analysis of HCoV-NH. Sequence analysis was
performed on the amplicon derived from each positive specimen (GenBank accession numbers AY870943AY871008). Phylogenetic analysis of a 126-bp portion of the HCoV-NHspecific primer region, which contains representative sequences of
HCoV-NH, is shown in figure 3. Overall, the amplified region
of the putative replicase 1a gene of HCoV-NH was highly conserved. These sequences closely matched the sequences of the
replicase 1a gene of the novel HCoV recently identified in The
Netherlands [10, 11]. Several distinct polymorphisms were identified among the HCoV-NH isolates; these polymorphisms were
not present in the HCoV identified in The Netherlands. Most
of the polymorphisms in the replicase 1a gene among HCoVNH isolates did not change the predicted amino acid sequence
(data not shown).
DISCUSSION
Using a PCR-based approach to screen for previously unknown
HCoVs, we identified evidence of the existence of a novel
HCoV, and we have described the clinical features associated
with infection with this agent. Our approach consisted of identifying conserved regions of the replicase 1a gene of a variety
of animal CoVs and HCoVs and then exploiting the conserved
genomic domains to identify a previously unknown HCoV. This
approach is based entirely on knowledge of CoV genomes, and
so identification of a CoV does not require the ability of the
virus to propagate in vitro or in animal models. This type of
genome-based screening has been used previously to identify
human bacterial pathogens by targeting conserved elements of
rRNA [14]. After the observation of CoV-like particles in reNovel Human Coronavirus JID 2005:191 (15 February) 495

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

0
3
9
2
6
9
3
1
0
5/7

Total
(n p 67)

spiratory secretions from individuals with SARS, broadly reactive PCR was used to identify the pathogen; unlike in the
present study, in which the replicase 1a gene was targeted, the
replicase 1b gene was targeted in the identification of the agent
that causes SARS [5].
In the present study, the initial set of primers, used to screen
pooled specimens, were less sensitive than the HCoV-NHspecific primers. Many of the respiratory specimens in the initial pool of 601 specimens that tested negative by PCR with
the replicase 1a gene primers tested positive by PCR with the
HCoV-NHspecific primers. Subsequent sequence analysis of
the target sites for the replicase 1a gene of HCoV-NH (data
not shown) revealed several mismatches between the genome
and primer sequence. This finding likely accounts for the relative decreased sensitivity of the replicase 1a gene primers.
Among the 601 specimens initially screened, the 2 pooled reactions that tested positive for HCoV-NH contained multiple
positive specimens. This may account for why these pooled
specimens tested positive. It is also unclear why primers constructed for the detection of the 2 common HCoVs, HCoV229E and HCoV-OC43, would not have detected HCoV-NH.
Presumably, the genomic sequence targeted for the detection
of HCoV-229E and HCoV-OC43 is specific for these viruses.
The sensitivity of the CoV consensus primers and the HCoVNHspecific primers is unknown. Determination of the sensitivity will require propagation of the virus and development
of a method for quantitating the number of viral particles (such
as an infectivity assay or a method based on electron microscopy, neither of which have been completely developed for
496 JID 2005:191 (15 February) Esper et al.

HCoV-NH). As further genetic data become available, moresensitive primers may be developed.
Our approach complements that of traditional virology, in
which viruses are identified by their effects on cells in tissue
culture. However, the identification of previously unknown viruses on the basis of in vitro findings has significant shortcomings, not the least of which is the inability to amplify and
characterize unique genetic sequences. This obstacle was overcome by Fouchier et al. [10] and van der Hoek et al. [11], who
used the tools of molecular biology to identify and sequence
the genome of a novel HCoV in The Netherlands after propagation in cell culture. hMPV was discovered by use of a similar approach that was based on in vitro virus propagation
and genomic amplification techniques [15]. Like the 2 groups
from The Netherlands, we have evidence that this novel HCoV
can be propagated in cell culture. The results of RT-PCR assays
of passaged cell culture supernatants suggested the growth of
HCoV-NH in vitro (data not shown).
There were early hints that HCoV-229E and HCoV-OC43 were
not the only common HCoVs. The first HCoV isolated, B814,
was not serologically related to either HCoV-229E or HCoVOC43 [16]. McIntosh et al. demonstrated that 3 of the 6 OC
strains isolated from adults with colds appeared to be serologically unrelated, or at least distantly related, to HCoV-OC43 and
HCoV-229E [17]. Experimental infection of human volunteers
with these 6 OC strains confirmed this finding [18]. It was,
therefore, unlikely that HCoV-229E and HCoV-OC43 were the
only HCoVs. The viral pathogen newly identified in New Haven

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

Figure 2. Weekly distribution of New Haven coronavirus (HCoV-NH)positive children !5 years old from January 2002 to mid-February 2003. For
clarity, only the first day of every other week is labeled. HCoV-NH infection in children hospitalized since birth at the newborn intensive care unit
(NICU), YaleNew Haven Hospital (CT), is represented by black bars.

Table 2. Respiratory specimens from children !5 years old

tested for New Haven coronavirus (HCoV-NH) from January 2002
to mid-February 2003.
No. tested
Year, month

Children

148
117
123
74
48
19
42
29
38
89
137
144

99
81
98
51
32
15
17
19
22
61
105
104

8
27
8
3
4
4
0
5
4
0
0
4

177
80
1265

129
62
895

4 (3.1)
8 (12.9)
79 (8.8)

(8.1)
(33.3)
(8.2)
(5.9)
(12.5)
(26.7)
(0)
(26.3)
(18.2)
(0)
(0)
(3.8)

and The Netherlands may represent the first of many antigenically distinct groups of HCoVs to be characterized.
In the present study, HCoV-NH was associated with both
upper and lower respiratory tract disease in infants and young
children. Because ours was not a population-based study and
because a control group was not included, it is impossible to
determine the spectrum of disease caused by HCoV-NH infection. Our study screened specimens submitted to a diagnostic laboratory only; therefore, the proportion of children
with clinical illness reported here likely does not reflect that of
the general population. Likewise, the issue of causality remains
unresolved. Prospective population-based studies are required.
Nonetheless, the identification of genetic sequences of this HCoV

Figure 3. Phylogenetic analysis of New Haven coronavirus (HCoV-NH). To construct the phylogenetic tree, sequences of a 126-bp portion of the
replicase 1a gene of a representative sample of HCoV-NH (NH) amplicons, the HCoV recently identified in The Netherlands (NL-63), HCoV-229E, and
transmissible gastroenteritis virus (TGEV) were used.
Novel Human Coronavirus JID 2005:191 (15 February) 497

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

2002
January
February
March
April
May
June
July
August
September
October
November
December
2003
January
February
Total

No. (%) of HCoV-NH

positive children

Specimens

from a large number of children with respiratory tract illnesses

strongly suggests that HCoV-NH infection plays a role in disease. Our data suggest that HCoV-NH, like other HCoVs, may
have a seasonal distribution [8]. Animal CoVs cause disease in
organ systems other than the respiratory tract, such as the
gastrointestinal system and the central nervous system [19].
SARS-CoV is excreted in the stool of infected individuals [6].
Therefore, it is necessary to determine whether HCoV-NH is
present in other body fluids and whether it is associated with
other clinical syndromes.
The percentage of children who tested positive for this novel
HCoV in our study (8.8%) was higher than the percentages of
patients who tested positive in the studies by van der Hoek et
al. (1.6%) and Fouchier et al. (2.9%) [10, 11]. The reasons for
this are unclear. Both our study and those in The Netherlands
screened respiratory specimens submitted to a diagnostic virology laboratory and used a PCRbased approach for screening. The characteristics of the study populations and the time
of the acquisition of the specimens screened may explain the
differences in the observed percentages of positive subjects between our study and the 2 studies in The Netherlands. Fouchier
et al. limited screening for the novel HCoV to 3 months during
late autumn and early winter [10] and, therefore, may have
missed the peak circulation of the virus in the population. Also,
we chose to screen only children !5 years old, whereas van der
Hoek et al. chose to screen individuals of all ages [11]. Similar
to other respiratory viruses, symptomatic infection with this
newly discovered HCoV may be more common during childhood; therefore, we may have been more likely to detect a higher
percentage of infected subjects in the present study. HCoVs can
cause severe disease in infants [20]. In the present study, a majority (50/79 [63.3%]) of HCoV-NHpositive children were, in
fact, !1 year old, and 34.2% were 3 months old.
Eleven (13.9%) of the 79 HCoV-NHpositive children in the
present study had been hospitalized since birth at the NICU.

Acknowledgments
We are indebted to George Miller, John F. Enders Professor of Pediatric
Infectious Diseases (Yale University School of Medicine), for his continued
support, intellectual and scientific input and exchange, and critical review
of the data and manuscript. We thank the staff of the Clinical Virology
Laboratory, YaleNew Haven Hospital (CT), for their assistance. We are
also indebted to Eugene D. Shapiro (Yale University School of Medicine),
for his thoughtful review of the manuscript.

References
1. Murray CJL, Lopez AD, Mathers CD, Stein C. The global burden of
disease 2000 project: aims, methods and data sources. In: Global programme on evidence for health policy. Geneva: World Health Organization, 2001.
2. Davies HD, Matlow A, Petric M, Glazier R, Wang EE. Prospective
comparative study of viral, bacterial and atypical organisms identified
in pneumonia and bronchiolitis in hospitalized Canadian infants. Pediatr Infect Dis J 1996; 15:3715.
3. Ruiz M, Ewig S, Marcos MA, et al. Etiology of community-acquired
pneumonia: impact of age, comorbidity, and severity. Am J Resp Crit
Care Med 1999; 160:397405.
4. Wright AL, Taussig LM, Ray CG, Harrison HR, Holberg CJ. The Tucson

498 JID 2005:191 (15 February) Esper et al.

5.

6.

7.

8.

9.
10.

11.
12.

13.

14.

15.

16.
17.

18.

19.

20.

Childrens Respiratory Study. II. Lower respiratory tract illness in the

first year of life. Am J Epidemiol 1989; 129:123246.
Ksiazek TG, Erdman D, Goldsmith CS, et al. A novel coronavirus
associated with severe acute respiratory syndrome. N Engl J Med 2003;
348:195366.
Drosten C, Gunther S, Preiser W, et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl
J Med 2003; 348:196776.
Rota PA, Oberste MS, Monroe SS, et al. Characterization of a novel
coronavirus associated with severe acute respiratory syndrome. Science
2003; 300:13949.
McIntosh K. Coronaviruses. In: Fields BN, Knipe DM, Howley PM,
eds. Fields virology. 3rd ed. Vol 1. Philadelphia: Lippincott-Raven, 1996:
1095103.
Pene F, Merlat A, Vabret A, et al. Coronavirus 229Erelated pneumonia
in immunocompromised patients. Clin Infect Dis 2003; 37:92932.
Fouchier RA, Hartwig NG, Bestebroer TM, et al. A previously undescribed coronavirus associated with respiratory disease in humans. Proc
Natl Acad Sci USA 2004; 101:62126.
van der Hoek L, Pyrc K, Jebbink MF, et al. Identification of a new
human coronavirus. Nat Med 2004; 10:36873.
Esper F, Martinello RA, Boucher D, et al. A 1-year experience with
human metapneumovirus in children aged !5 years. J Infect Dis 2004;
189:138896.
Esper F, Boucher D, Weibel C, Martinello RA, Kahn JS. Human metapneumovirus infection in the United States: clinical manifestations
associated with a newly emerging respiratory infection in children.
Pediatrics 2003; 111:140710.
Relman DA, Schmidt TM, MacDermott RP, Falkow S. Identification
of the uncultured bacillus of Whipples disease. N Engl J Med 1992;
327:293301.
van den Hoogen BG, de Jong JC, Groen J, et al. A newly discovered
human pneumovirus isolated from young children with respiratory
tract disease. Nat Med 2001; 7:71924.
Bradburne AF. Antigenic relationships amongst coronaviruses. Arch
Gesamte Virusforsch 1970; 31:35264.
McIntosh K, Kapikian AZ, Hardison KA, Hartley JW, Chanock RM.
Antigenic relationships among the coronaviruses of man and between
human and animal coronaviruses. J Immunol 1969; 102:110918.
Bradburne AF, Somerset BA. Coronative antibody tires in sera of
healthy adults and experimentally infected volunteers. J Hyg (Lond)
1972; 70:23544.
Holmes KV, Lai MM. Coronaviridae: the viruses and their replication.
In: Fields BN, Knipe DM, Howley PM, eds. Fields virology. 3rd ed.
Vol 1. Philadelphia: Lippincott-Raven, 1996:107593.
McIntosh K, Chao RK, Krause HE, Wasil R, Mocega HE, Mufson MA.
Coronavirus infection in acute lower respiratory tract disease of infants.
J Infect Dis 1974; 130:5027.

Downloaded from http://jid.oxfordjournals.org/ by guest on October 20, 2016

There were 2 apparent temporal clusters of HCoV-NH infection

at the NICU, during February 2002 and during January and
February 2003. Although these may represent nosocomial outbreaks, other modes of infection, such as perinatal transmission,
cannot be discounted. HCoV-NH was detected in a child diagnosed with hydrops fetalis. Whether this represents congenital infection with HCoV-NH remains to be determined.
In conclusion, the present results demonstrate the power of
the tools of molecular biology to define and characterize potential infectious agents associated with human disease. The
novel HCoV identified in New Haven and The Netherlands
was associated with both upper and lower respiratory tract
disease in infants and young children. Whether this virus is
associated with other clinical syndromes remains to be determined. Population-based studies are required to define the burden of disease caused by this novel HCoV, and such studies
could provide information on causality.